Abstract Previous evidence has shown that perceptual cues, such as weight, can affect metacognitive monitoring and control, providing evidence for the embodied nature of metacognition. Moreover, some researchers believe that the influence of perceptual information on metacognition is automatic and unconscious. However，there is no direct evidence to support this hypothesis. According to Reder’ s view, metacognitive monitoring and control can occur without conscious awareness. For example, if people are unaware of what causes them to select one strategy rather than another, then the strategy select is an unconscious metacognitive control. Therefore, it is critical to reveal people’s basis of the metacognitive monitoring and control establishing whether the influence of weight on metacognition is unconscious. For this purpose, we adopted the method of structural knowledge attribution. This method requires participants to report their basis of judgment or decision in each trial using one of a set of fixed options: guess, intuition, memory, and rules.“Guess” indicated that the judgment or decision was based on nothing at all; ‘‘Intuition” indicated that the judgment or decision was based on a hunch or feeling that could not be articulated; ‘‘Memory” indicated that the judgment or decision was based on a recollection; ‘‘Rules” indicated that the judgment or decision was based on a rule (or a belief, or a theory) that could be stated if asked. The “guess” and “intuition” responses were taken to indicate that the basis of the judgments or decisions were unconscious because people had no idea why they made such a judgment or decision. The “memory” and “rules” responses were taken to indicate that the basis of the judgments or decisions were conscious because people know why they made such a judgment or decision. Therefore, the metacognitive process is unconscious when monitoring or control was based on guessing or intuition. And conversely，the metacognitive process is conscious when monitoring or control was based on memory or rules. Following this logic, if the influence of weight on metacognitive monitoring and control is unconscious, the effect of weight on monitoring and control would occur when monitoring and control were based on guessing or intuition. In two experiments, participants studied word pairs affixed to heavy or light pillow boxes, made judgments of learning (experiment 1) or selections for items (experiment 2), as well as reported the basis of each decision. Results showed: 1) the effect of weight on judgments of learning only occurred when the judgments were made based on guessing, JOLs were higher for items affixed to the heavy boxes than those affixed to the light boxes; 2)the effect of weight on selections of items only occurred when the selections were made based on guessing, participants preferred to first select word pairs affixed to heavy paper-boxes, compared with those to light paper-boxes. These results indicate that the influence of weight on metacognitive monitoring and control is unconscious.